the acceptance of wearable devices for personal...

The acceptance of wearable devices for personal healthcare in China

University of Oulu

Department of Information Processing

Science

Master’s Thesis

Min Weng (2442857)

Date 17/04/16

2

Abstract

Context: In recent years, health and fitness have drawn greater attention to consumers

in China. The demand of wearable devices has risen and the number of potential

customers is large. This study would like to explore if the wearable devices match

customers’ desire and expectation, and what influence users’ behavioural intention to

use wearable devices.

Aim: This thesis aims to examine the acceptance of wearable devices, in particular,

smart bands and dedicated healthcare applications, in order to find usage patterns,

preferences with regard to product features, and the determining factors of users’

acceptance.

Method: This study is a descriptive and explanatory research. First, a literature

review on 1) wearable devices in healthcare, and 2) technology acceptance model and

related models was conducted. Then, a research model with 11 hypothesizes was

derived based on the technology acceptance model (TAM), unified theory of

acceptance and use of technology (UTAUT), trust model, and technological

personality construct. The research model and the hypotheses were tested by

conducting a quantitative questionnaire survey in China. 158 responses were analysed

using Partial Least Squares Structural Equation Modelling in smartPLS software

package.

Results: The factors affecting directly the user’ intention to use smart bands are:

perceived usefulness, social influence, affinity, and compatibility. Trust and perceived

ease of use affect behavioral intention indirectly through perceived usefulness.

Keywords

wearable devices, technology acceptance model, unified theory of acceptance and use

of technology, smart bands, healthcare, structural equation modelling

Supervisor

Dorina Rajanen

3

Foreword

I started working on this thesis at the end of May, 2015. The initial aim was to

research on topics related to healthcare. After several discussions and brainstorming,

my supervisor, Dorina Rajanen, helped me to make sure the orientation. I would like

to express my gratitude to her for the useful comments, suggestions and engagement

through the learning process of this master thesis. I also cannot deny her

encouragements that pushed me forward made me benefit a lot. And looking back at

my first steps, I realize I have made significant progresses.

I would like to thank the participants in my questionnaire, who have willingly shared

their precious time. Without them, I cannot have the indispensable data for

researching.

Last but not the least, I would like to thank my parents for supporting me spiritually

throughout writing this thesis and my life in general.

Min Weng

Oulu, April 17, 2016

4

Contents

Abstract .......................................................................................................................... 2

Foreword ........................................................................................................................ 3

1. Introduction ............................................................................................................ 6

2. Problem statement, study aim, and research question .......................................... 8

3. Literature review ................................................................................................... 10

3.1 Wearable devices ........................................................................................... 10

3.2 Wearable devices in healthcare ..................................................................... 11

3.2.1 Smart bands features .............................................................................. 11

3.2.2 Current situation and future prospect of healthcare wearables ............ 13

3.3 Applications for wearable devices in healthcare ........................................... 13

3.3.1 Application functions .............................................................................. 14

3.3.2 Current situation and future prospect of smart bands apps .................. 15

3.4 User evaluation of wearable devices and apps ............................................. 15

3.4.1 User characteristics ...................................................................................... 16

3.4.2 Privacy .......................................................................................................... 16

3.4.3 Usability ....................................................................................................... 17

3.4.4 Adoption of new technology ....................................................................... 17

4. Theories of technology acceptance ...................................................................... 19

4.1 Technology Acceptance model (TAM) ............................................................ 19

4.2 Technology Accepted Model 2 (TAM2) .......................................................... 20

4.3 Unified Theory of Acceptance and Use of Technology (UTAUT) .................... 20

4.4 Trust and TAM ................................................................................................. 21

4.5 Technological personality ................................................................................ 22

5. Research model and hypotheses .......................................................................... 23

5.1 PU, PEOU, and Trust ....................................................................................... 26

5.2 Social influence .............................................................................................. 26

5.3 Technological personality ............................................................................... 27

5.4 Moderators .................................................................................................... 28

6. Research methodology ......................................................................................... 30

6.1 Quantitative research approach .................................................................... 30

6.2 Questionnaire development .......................................................................... 30

6.3 Pilot study ...................................................................................................... 33

6.4 Data collection ............................................................................................... 33

6.4.1. Sampling .................................................................................................. 33

5

6.4.2. Questionnaire administration ................................................................. 34

6.5 Data analysis .................................................................................................. 34

7. Results ................................................................................................................... 36

7.1 Sample characteristics ................................................................................... 36

7.2 Smart bands usage patterns .......................................................................... 37

7.3 Evaluation of measurement model ................................................................ 40

7.4 Evaluation of structural model ....................................................................... 42

8. Discussion.............................................................................................................. 46

8.1 Implications of results .................................................................................... 46

8.2 Limitations of studies and future work .......................................................... 47

9. Conclusion ............................................................................................................. 49

Reference ..................................................................................................................... 50

Appendix. Questionnaire ............................................................................................. 59

6

1. Introduction

As the attention to healthy living and healthy lifestyle arises, wearable devices and

dedicated healthcare applications become more and more popular. The most familiar

wearable devices today are smart bands, glasses and watches (Morris, 2015). Morris

(2015) also conceptualized wearable devices as electronics or computers that can be

worn on the body when inserted into items of clothing and accessories.

The wearable devices combine with web and mobile apps to achieve various

functionalities including sports tracking, sleep monitoring, event notifications, and

heart rate recording. Thus, wearable devices are not only a kind of hardware, but

realize their powerful features through web and mobile application software that

interacts with the data collected by the devices. Wearable devices are used for data

collection and monitoring, while dedicated healthcare applications are employed for

data processing and exporting.

Morris (2015) marked the milestones of wearable technologies according to

Knoblauch (2014)’s study. The first conceptualization of wearable computing dates

from late 1990s, thus this technology is not relatively new. Then in 1970s, wearable

computer developed to increase the chance of winning at blackjack (Knoblauch,

2014). A calculator wristwatch was produced as the first consumer device in the 1980s

(Morris, 2015). Then in 2009, FitBit’s first wearable activity-tracking device was

launched. And more recently in 2013, Google Glass was released (Knoblauch, 2014).

Mann (2001) called it wearability (or portability) as devices that are attached to or

carried by body, starting from devices being installed in frame, being hand held, to

being wearable or implanted.

However, there are still many challenges regarding the development and use of this

emerging technology. First of all, generally speaking, accurate measurement requires

a perfect touch with the skin, which might lead to users’ discomfort (Knight,

Deen-Williams, Arvanitis, Baber, Sotiriou, Anastopoulou, & Gargalakos, 2006).

Moreover, for an example, a regular rotation of an arm can be mistakenly logged in

the database as a walking motion event. Secondly, although many wearable devices

can store data offline, it is still challenging to display the result by the mobile apps.

Thirdly, the users’ acceptance of wearable devices seems to be also challenging. As an

illustration, according to a study in 2014 in China, 40.5% of respondents held the

opinion that the functionality of wearable devices and dedicated healthcare

applications are incomplete (Tencent ISUX, 2014). Furthermore, the same study

showed that 45.7% of users abandoned using smart bands within a month.

7

Lymberis mentioned that the wearable devices in the fields of healthcare research and

industrial community are facing a few of common key issues, although they focused

on specific systems and applications to satisfy users (biomedical sensors, data security

and confidentiality, medical knowledge basement, user acceptance and awareness, for

instance) (Lymberis, 2003). Nevertheless, after a period of years, wearable devices

tends to be more handiness, fashionable, and lightweight. (Wright & Keith, 2014).

In recent years, health and fitness have drawn great attention to consumers in China.

The demand of wearable devices has risen and the number of potential customers is

large. Rely on the advantages in health care delivery, wearable devices are primarily

used in health care industry (Wu, Li, Lin, & Goh, 2015).

This study examines the adoption of the wearable devices and dedicated healthcare

applications for wearable devices in order to find usage patterns and preferences with

regard to product features, as well as determinants of users’ acceptance.

The thesis is organized as follows. Chapter 2 presents the problem statement, aim of

the study and the research question. Chapter 3 introduces the wearable devices in

healthcare, smart bands and related applications. Chapter 4 presents the theoretical

background of the research, namely, the technology acceptance model (TAM) and its

variants TAM2 and the unified theory of acceptance and use of technology model

(UTAUT). In addition, the trust model and technological personality are described.

Chapter 5 describes the proposed research model and hypotheses. Chapter 6 describes

the research methodology for data collection and analysis, and Chapter 7 presents the

data analysis results. Chapter 8 discusses the implementation of the research,

limitations of the study, and future work proposals. Finally, Chapter 9 concludes the

thesis study.

8

2. Problem statement, study aim, and research

question

Wearable devices are one of the initial products developed in the areas of healthcare

and mobile intelligence (Salah, MacIntosh, & Rajakulendran, 2014). Nowadays, they

can monitor real-time indicators of the human body, such as heart rate, muscle

activities, etc. In the future, some measurements which are currently available only in

hospitals (for example, brain waves, and blood cells exams) will gradually move to

intelligent terminals such as wearable devices. With the development of technology,

big data, such as personal genome, living habits, environmental factors, and even

real-time heart rate, and blood pressure, could be monitored and analyzed in the future

so that early warning of disease will come true (Marcengo & Rapp, 2014).

However, besides the current technological challenges to achieve this objective, there

are still users’ issues that have to be investigated. Does the wearable technology

match potential consumers’ desires and expectation? According to some authors,

patients’ needs of e-health services are clear and definite, such as they desire online

services by their own health care provider (Wilson & Lankton, 2004). Nevertheless, it

is not so clear whether the available e-health services are what the people desire and

this mismatch between the needs and offers may impede user’s acceptance of new

e-health or mobile health technology. In this study, we focus our investigation upon

the wearable devices and dedicated mobile applications in healthcare.

The main research question of this study is: What is the state of acceptance of

wearable technology for personal healthcare in China? Through a questionnaire

survey in social media, the aim of this thesis is to provide a state-of-the-art of the

acceptance and adoption of wearable technology (both devices and dedicated

applications) for personal healthcare in China. Among the topics that this study is

aiming to uncover are: the extent of user adoption, the extent of user acceptance,

usage patterns and preferences with regard to device types, application types, and

frequency of use, as well as the determinants of users’ acceptance of wearable

devices.

In this thesis, adoption means the actual use (past or current) of wearable devices and

dedicated applications. This follows the definition in Table 5.1. On the other hand,

acceptance is defined as intention to use or behavioral intention (Davis, 1989;

Venkatesh and Davis, 2000; Venkatesh et al., 2003; Gefen, Karahanna, Straub, 2003;

Compeau & Higgins, 1995; Aldaz et al. 2009; Rubin, 1981; Rogers, 1995).

9

The study goals are thus descriptive and explanatory. In the descriptive part, we aim

to uncover the extent of adoption and acceptance of wearable devices in China as well

as usage patterns; in the explanatory research, we aim to shed light on the factors

affecting acceptance or behavioral intention to use wearable devices for healthcare.

The availability of medical resources is usually insufficient, especially in developing

countries, and China is one of them. The adoption of wearable devices could ease the

stress of public medical service. The information on usage patterns and the

determining factors of acceptance among potential customers can give developers a

feedback for the future improvement of wearable devices. Based on the research of

users’ adoption towards wearable devices in healthcare can help related companies

have its needs-based positioning and make special marking strategies to reach their

marketing target.

In the study, the acceptance of wearable devices (technology) covers the acceptance

of wearable devices and of related applications. The study focuses on smart bands for

healthcare and, thus, a questionnaire have been created for smart bands acceptance

investigation. For the descriptive research goal, the data collected are analyzed using

descriptive statistics such as central tendency measures (mean), variation measures

(frequency tables, standard deviation), and measures of correlation presented in tables

and charts. For the explanatory part of research, the determining factors are obtained

using the structural equation modeling.

10

3. Literature review

3.1 Wearable devices

The evolution of digital and mobile technology has transformed our lives to be digital

lives. People’s desire are not only to carry digital products, but to let electronics or

computers be part of their bodies. Then wearable technology or known as “wearable

devices” emerged.

Wearable devices (a.k.a. wearable technologies) are defined by Tehrani & Michael

(2014) as computers that reside on clothes and/or accessories (such as eyeglasses,

rings) that are comfortable to wear. This combination of features determined that user

performance in some areas, such as aircraft maintenance, navigational assistance, and

vehicle inspection, has dramatically reformed (Billinghurst & Starner, 1999).

The first wearable devices originated in the 1960s for cheating in casinos. Gamblers

usually count cards or improve their odds at the roulette table (Morris, 2015). From

gaming and entertainment to practical purpose, wearable devices tend to improve

enjoyment and quality of lifestyle. For example, smart motorcycle helmets are

developed to help observe surrounding traffic situations (Morris, 2015). Certainly,

wearable technology reinvest normal products with fresh digital function as well as

basic functionality. Digital clothes are designed for performance at night and rings as

a notification hub for delivering message, receiving incoming calls and push

notifications (Morris, 2015). Recently, smart bands, eyeglasses and wristwatches have

become available to consumers and become popular (Morris, 2015).

Providing functionality of perception and scanning, wearable technologies tend to be

more intricate than normal handheld technologies (Billinghurst & Starner, 1999). And

also tend to be more sophisticated as to receive bio-feedback (Punagin & Arya, 2015).

However, the main purpose of wearable devices is to simplify people’s daily life and

free up the arms (Robert, 2015). Previously, wearable devices mainly used in the field

of military (Tehrani & Michael, 2014). Nowadays, combined with aesthetics and

fashion design, demand of wearable devices continues to grow in various fields

(Tehrani & Michael, 2014). And they have more influence on finance, education,

gaming, and enterprise, and especially in healthcare and medicine.

11

3.2 Wearable devices in healthcare

Safavi & Shukur mentioned that healthcare is closely related to our daily life, so there

is no doubt why it attracts attentions in wearable devices (Safavi & Shukur, 2014).

The emerging of advanced communications technology make smart phone entirely

possible to achieve the goal of display third-party applications and put it into use

(Mackert, Champlin, Holton, Muñoz, & Damásio, 2014). Certainly, most individuals

own smartphones. One of the concerns now of healthcare and technology providers is

to find the ways of connecting that to benefit from this trend.

In the field of healthcare, a full and detailed report provided by wearable devices

helps users to keep monitoring a healthier life (Sun & Rau, 2015). For example, sign

language rings can translate hand movements into spoken words, and vice versa, and

thus help the hearing-impaired users to communicate more easily with others and in

this way, disabled people could change their life pattern, which is of great significance

(Michael, 2013).

In the field of healthcare, examples of wearable devices comprise smart bands,

watches, glasses, and other accessories (Robert, 2015). Most of them have the

function of health monitoring, physiological activities tracking, notification, sleep

monitoring, and heart rate recording (Robert, 2015). With the progress of wearable

technology and the transformation of user requirements, wearable devices and its

relevant applications are also constantly changing.

Smart bands (or wearable bracelets) is one kind of wearable devices that record

real-time data such as for monitoring physical activity and sleeping patterns (Magno,

Porcarelli, Brunelli & Benini, 2014). These data synchronize with mobile phone or

tablet with the purpose to guide a healthy life through data. In the next section, the

features of smart bands are described.

3.2.1 Smart bands features

Smart bands are characterized by a series of features that can be classified into three

main categories: 1) appearance (that is, look and feel of the smart band); 2)

functionality (defined here by information (measures) they provide as well as by the

operation of the device by the user); and 3) technical requirements for operation (such

as operation system platform or environmental requirements).

Among functionality features, Safavi and Shukur (2014) mentioned in their study the

tracking of user’s position, by means of accurate position indicator and sensors inside

the wearable devices. Moreover, steps can be recorded, even heart rate measured by

12

the built-in sensor. They also point out that these devices are incapable of recording

and transmitting data until internet connected. Table 3.1 shows the functionality

features of common or popular smart bands devices for health care.

Table 3.1. Functionality features of smart bands

Features Description References

Measures Steps

Calories Burned

Distance

Sleep Quality and Duration

(Misfit, 2015)

Sensors and Components GPS

3-axis accelerometers and gyroscope,

Digital compass

Optical heart rate monitor

Altimeter

Ambient light sensor

Vibration motor

Mio continuous optical heart rate

(The NPD Group, 2015)

(micoach, 2015)

Appearance of wearable devices usually attract consumers at first glance (Table 3.2).

Nowadays, it become more and more colorful, with great fashion sense, small and

convenient.

Table 3.2. Appearance of smart bands

Appearance Multicolour

Shape (watch or bracelet).

Tehrani, Kiana and

Andrew (2014)

Material Aluminium, elastomer material or silicon

strap.


(Misfit, 2015)

(micoach, 2015)

Weight For example, 9.4g (Misfit Shine) (Misfit, 2015)

Display LED or LCD

Touch screen

Backlight for low light visibility


Battery or charge Button cell, Lithium-polymer or USB

charge


Water proof or splash proof Rain-proof

Sweat-proof from a hard workout

Wet-proof in the shower.

(Jawbone, 2015)

Regarding the technical requirements, as seen above, one condition is to have Internet

connection. Other requirements are regarding the environment temperature, operating

latitude, power source, as well as operating system platform. (Table 3.3)

13

Table 3.3. Technical requirements of smart bands

Environmental Requirements Temperature: -4° to 113°F

Operating altitude: less than 30,000 feet


Compatibility Ios

Android

Windows.


(Misfit, 2015)

(micoach, 2015)

Sync method Bluetooth 4.0 or 4.1

Bluetooth Low Energy

(Misfit, 2015)

3.2.2 Current situation and future prospect of healthcare wearables

Currently, wearable devices only collect some superficial health data, next step is to

deepen the data analysis, attract consumers to upload these data and their health and

daily life together, and to encourage young consumers to care about their health data

(Punagin & Arya, 2015).

Wearables have the potential to offer a telemedicine platform to offer better quality

and lower cost for the patient and citizen. (Lymberis, 2003). Moreover, wearable

devices research and development should emphasize people with chronic diseases so

as to expand the market better (Hung, Zhang, & Tai, 2004).

When the main task of wearable devices shift from basic healthcare to medical

treatment, the accuracy of measurements need to be improved, as well as the

sufficient privacy capabilities related to service. With the development of technology,

the prospect is that the wearable devices become more comfortable and friendly, even

be part of our bodies (Anderson & Rainie, 2014).

When get out of the limitation of wearable devices in the field of healthcare, it will

have several ways for expanding. Hemmadi (2014) points out that businesses are

enthralled with wearable devices especially for office use. In Hemmadi’s study,

according to the technology market intelligence firm ABI Research, by the year of

2019, the whole market of enterprise wearable devices and applications could reach

US$18 billion. Credit Suisse also issued a report that under the drive of Apple and

Google, wearable technology market will grow to US$30 billion - 50 billion in the

future.

3.3 Applications for wearable devices in healthcare

Applications for wearable devices in healthcare are defined as any applications or

technology that are designed to interact with consumers directly through the platform

of computer, smartphones or tablets, no matter if they have the professional

knowledge of healthcare, and that help a patient better monitor his or her healthcare

14

through individualized information (Parsons, 2009).

Mobile health or m-health is regard as an important field due to the many benefits of

health apps (Flaherty, 2014). The field of mobile health has enabled physical

examination to be conducted via wearable devices (Safavi & Shukur, 2014).

The most famous applications for smart bands in healthcare are Endomondo,

Runtastic, Wahoo Fitness, Runkeeper, miCoach, Nike+ running, Fitbit, Misfit,

Jawbone up, etc.

3.3.1 Application functions

Applications represent the platform for the wearable devices to display their recording,

monitoring, and data analysis. The two main functions are 1) see progress, and 2)

record workouts.

Presently, the feature of music play and social sharing become indispensable.

Although these are not the key features of wearable devices, users have the desire to

enjoy in the atmosphere of activities and then interact with relatives and friends for

experience sharing. It is also a smart way for promotion.

Table 3.2. Functions of smart bands applications

Function Description References

GPS tracking

Live Map

Track activities including steps, distance, calories

burned, intensity, floor climbed, streaks, milestones

and active minutes. Recognize running, cycling,

walking, mountain biking, kayaking, skiing, floor

climbing and other exercises.

(Hunder armour

connected fitness,

2015)

(Jawbone, 2015)

Goal coach Get coach feedback on performance for each mile or

kilometre. Setting with a challenging but realistic goal.

Stay focused by setting step, weight and activity goals.

Achieve weight goals with Calorie Coaching.

(Hunder armour

connected fitness,

2015)

(FitnessKeeper,

2015)

(Fitbit, 2015)

See progress

See workout history

View every step of the workout progress with punch

card and read charts, graphs, tables that including split

times, workout map, friends and comments. Diversity

of analysis strategies. Table, graph, etc.

(The NPD Group,

2015)

(Hunder armour

connected fitness,

2015)

Sleep monitoring Monitor sleep quality automatically and set a silent

alarm. Can accurately detect fall asleep and wake up

time, estimate time of deep sleep, and snoring.

(Misfit, 2015)

(The NPD Group,

2015)

15

Heart rate recording Wrist-based heart rate, continuous and automatic. (Hunder armour

connected fitness,

2015)

Food log Monitor daily food intake. (Misfit, 2015)

Track weight Create a goal, and track weight. (Misfit, 2015)

Notification and

reminders

Notice of the daily schedule, work schedule, exercise,

and so on. Including phone, MSG.

(Fitbit, 2015)

Music play Control the play list in the mobile phone and play. Get

the perfect playlist built with music you love.

(FitnessKeeper,

2015)

Social sharing Celebrate progress with friends.

Punch card and share with friends to social Apps.

(FitnessKeeper,

2015)

(Nike, 2015)

Earn badges and

reward

Reach milestones to get Badges.

Accumulate mileage and get reward or gifts.

(Fitbit, 2015)

(Runnit, 2015)

3.3.2 Current situation and future prospect of smart bands apps

There exist a variety of applications for wearable devices in healthcare to help build a

healthy life for individuals.

For example, Flaherty (2014) introduced a case where a woman with impaired vision

used a mobile application (namely, Portable Eye Examination Kit) for taking a photo

for her doctor to conduct a vision test and report diagnosis of cataracts.

These applications can be installed in a smartphone and facilitate telemedicine.

However, with convenient applications for wearable devices in healthcare there is the

issue of regulation. When the intended use of an application is for diagnosis,

treatment, or prevention of disease, the application is a device that needs to be

regulated (US Food & Drug Admin, 2013).

Moreover, privacy and security require more and more attention in the future. The

Trend Micro's vice president of technology and solutions – J.D. Sherry said in a

statement that most of consumers are not aware of regularly hack in smart appliances

(Eddy, 2014).

3.4 User evaluation of wearable devices and apps

User evaluation of e-health and m-health devices and apps address two aspects: 1) the

medical information provided by the technology, and 2) the interface and functionality.

Regarding the first aspect, a survey by Philips Electronics in year 2012 revealed that

11% of American citizens believe that it there were not access to health information

16

on website or inside the applications, they might already be dead or severely

incapacitated. It was reported in Philips’ survey that one fourth of respondents use

symptom checker websites as much as they visit the doctor, even another 27 percent

prefer interact with health applications than doctors (Royal Philips Electronics, 2012).

In the next sections user evaluation will be reviewed from four aspects: user

characteristics, privacy concerns, usability of wearable devices, and users’ adoption of

new technology.

3.4.1 User characteristics

In Gaul and Ziefle (2009)’s survey of acceptance for a medical stent implemented into

the body, results indicated that the middle-aged generation had the highest degree of

acceptance while the oldest generation showed the lowest acceptance level and

highest rating on potential barriers.

Analogously, wearable devices are suitable for various people from young to elders.

Especially the one who want to monitoring their work out and physical condition.

Most of them have the ability to buy and have the desire to buy new products. In this

study, users are defined as present users, past users, and potential users of smart bands.

For some potential users, recommended applications by devices companies are a way

to test devices to help them make a buying decision. Of course, they can try the

different applications.

3.4.2 Privacy

Users usually are prompted by some healthcare applications of inputting their

personal information which have already raised a number of privacy concerns. When

users input (sometimes very personal) information into a health app, such as weight,

snoring or not, heart rate, and daily workout progress, it is not clear where this

information goes and what the app developer or phone manufacturer does with it

(Flaherty, 2014).

As privacy is given more and more attention, an increasing number of companies

recommend users to correctly install and set up the devices and applications, and to

use a secure password for the connectivity to ensure privacy (Eddy, 2014). Moreover,

application updates and hardware updates both are essential in time. While wearable

devices and supported applications usually are for personal use (Eddy, 2014).

While the positive effects of wearable technology cannot be ignored, the issues related

17

to privacy continue to surface and mainly on the question of who may have the

authority to access the collected data. Not only businesses may benefit from these data

information, but also some research or experiments would rely on the personal

information. Samavi, Consens, & Chignell (2014) found out that although

study-respondent are aware of privacy concerns, 60% of them reported that they just

sign but not read privacy agreements in detail.

Wright & Keith (2014) pointed out that a wearable start-up designed some accessories

including bracelets, necklaces, and key chains, and one of the special features in these

accessories primarily designed is for personal security.

3.4.3 Usability

Nielsen (2012) define usability as being a quality attribute that assesses how easy user

interfaces are to use. Furthermore, usability is defined by five quality components and

can be defined as follows when linked to smart bands and its applications:

1. Learnability: How easy to use are the smart bands?

2. Efficiency: What is the performance of smart bands?

3. Memorability: How many customers will keep on using and recommend to others

smart bands?

4. Errors: What are the lack of functionality and the shortcomings of smart bands?

5. Satisfaction: mood or attitude of the users while using these devices.

There isn’t explicit definition from studies reporting of usability on wearable devices

or smart bands. Sultan (2015) comment on his study that the glorious future of

wearable technology is due to wearable devices could be enhanced connectivity,

improved usability, reduced cost, increased reliability and long battery life (Sultan,

2015).

3.4.4 Adoption of new technology

Sun & Rau (2015) show that users’ attitudes toward e-health technology include trust,

privacy concerns, intimacy facets, and compatibility with users’ needs. According to

the authors, ease of use, trust, and privacy concern impact the acceptance and

selection of apps.

PwC defined the millennial generation as the current 18 to 24 age, and they

discovered Millennials are double more likely than adults ages more than 35 to be

very willing to adopt smart bands if some company pays for it (PwC, 2014). While

most of smart bands and other wearable devices which already found in the market

are below expectations (PwC, 2014).

18

In order to find out the how is state-of-art of smart bands and its dedicated healthcare

application that is very representative of wearable devices in the field of healthcare,

this thesis addresses the acceptance and adoption of smart bands in China, by adapting

the existing models on technology acceptance and use and conducting an online

questionnaire survey for collecting and analyzing data on smart bands and healthcare

applications. The next chapters describe the technology acceptance models and the

research model, as well as the research methodology for data collection and analysis,

the results, and the discussion of the findings.

19

4. Theories of technology acceptance

4.1 Technology Acceptance model (TAM)

Research on technology acceptance models started in 1980s and one of the most

prominent model in this area was technology acceptance model (TAM) developed by

Davis (Davis, 1989). TAM was derived from the theory of reasoned action (TRA;

Ajzen & Fishnein, 1980). The model explains the acceptance of technology

through measuring individuals’ intentions to use a technology and determining factors.

Holden and Karsh (2010) point out that in order to promote technology acceptance

and even increase information technology use, knowing which of the factors

negatively influence technology acceptance would help organizations to better control

those factors.

TAM explains behavioral intention or acceptance by two important factors: perceived

ease of use (PEOU) and perceived usefulness (PU) (Figure 4.1). Perceived usefulness

is defined as “the degree to which a person believes that using a particular system

would enhance his or her job performance.” Perceived Ease of Use is defined as “the

degree to which a person believes that using a particular system would be free of

effort.” (Davis, 1989).

Figure 4.1. Technology Acceptance Model (Davis, 1989)

PU and PEOU influence the attitude towards use of technology (ATT) which in turn

influences the behavioral intention to use (BI). Moreover, PU has a direct effect on BI.

The model highlights also a causal relationship of PEOU on PU (perceived ease of

use is shown to affect perceived usefulness). In the model, BI or acceptance will lead

to actual use (AU).

20

4.2 Technology Accepted Model 2 (TAM2)

Venkatesh and Davis (2000) proposed an updated model, known as TAM2, which

added subjective norm (SN), image, job relevance, output quality, results

demonstrability as factors that influence PU. The model kept PEOU, but omitted ATT

due to weak predictors of either BI (Acceptance) or AU (see also Wu & Wang, 2005).

In TAM2, PU, PEOU, and SN have direct influence on BI which affects actual use

(AU) (Figure 4.2).

Figure 4.2. Technology Acceptance Model 2 (Venkatesh and Davis, 2000)

The following definitions describe the determining factors of PU that are specific to

the TAM2 (Venkatesh and Davis, 2000)

Subjective norm: “Person's perception that most people who are important to him

think he should or should not perform the behavior in question.”

Image: “The degree to which use of an innovation is perceived to enhance one's

status in one's social system.”

Job relevance: “Individual's perception regarding the degree to which the target

system is relevant to his or her job.”

Output quality: “The degree to which an individual believes that the system

performs his or her job tasks well.”

Results demonstrability: “Tangibility of the results of using the innovation.”

4.3 Unified Theory of Acceptance and Use of Technology

(UTAUT)

Few years later, Venkatesh and colleagues developed an improvement in TAM2,

namely, the unified theory of acceptance and use of technology that defines the earlier

concepts of PU, PEOU, and SN as performance expectancy, effort expectancy, and

social influence, respectively (Venkatesh, Morris, Davis, & Davis, 2003). UTAUT

explains use behavior by BI and facilitating conditions - a new component added into

21

the technology acceptance and use model (Figure 4.3). Early tests of UTAUT account

for 70% of the variance in usage intention and about 50% in actual use (Holden &

Karsh, 2010). In addition, four key moderating variables are included in the model:

experience, voluntariness, gender, and age. In Venkatesh and colleges’ research,

Significant moderating influences of experience, voluntariness, gender, and age were

confirmed as integral features of UTAUT.

Voluntariness of use means“the degree to which use of the innovation is perceived as

being voluntary, or of free will” (Moore and Benbasat, 1991)

Figure 4.3. Unified Theory of Acceptance and Use of Technology (Venkatesh et al., 2003)

Four determinants are defined as follows (Venkatesh et al., 2003).

Performance Expectancy: “the degree to which an individual believes that using

the system will help him or her to attain gains in job performance”.

Effort Expectancy: “the degree of ease associated with the use of the system”.

Social Influence: “the degree to which an individual perceives that important

others believe he or she should use the new system”.

Facilitating Conditions: “the degree to which an individual believes that an

organizational and technical infrastructure exists to support use of the system”.

4.4 Trust and TAM

Wu and Wang (2005) summarized in their study that connections between trust and

TAM have been widely discussed in literature especially the relationships between PU,

PEOU, and trust (Gefen et al., 2003; Pavlou, 2003; Saeed, Hwang, & Mun, 2003;

Gefen, 2004).

The contribution of trust is to make sure of better rewards from economic activities so

that people make efforts to reduce this social complexity and avoid risk from being

exploited (Wrightsman, 1972). Using wearable devices can be regard as economic

22

activities that’s why we try to include trust into the research model.

In particular, Gefen et al. (2003) gave a brief model of trust and TAM, which is

suitable to be part of our research model that will be discussed below.

4.5 Technological personality

Aldás-Manzano, Ruiz-Mafe, & Sanz-Blas (2009) had the findings that personality

variables (affinity to mobile telephones, compatibility and innovativeness) have a

direct and positive influence on the intention to engage in M-shopping. Similarly, it

can be included to evaluate behavioral intention to use wearable devices.

Aldás-Manzano et al. (2009) updated the definition as follows:

Innovativeness: “the degree of interest in trying a new concept, or an innovative

product or service” (Rogers, 1995).

Affinity: the perceived importance of the medium in the life of the individual

(Rubin, 1981).

Compatibility: “the degree to which an innovation is perceived as consistent with

the existing values, past experiences, and needs of potential adopters” (Rogers,

1995).

Koufaris (2002) found that perceived Web skills are positively related with shopping

enjoyment and concentration of online consumers. Similarly, mobile technology skills

can be added for testing behavioral intention of using wearable devices.

Mobile technology skills: “An individual judgement of one’s capability to use

mobile technology” (Compeau & Higgins, 1995).

23

5. Research model and hypotheses

To model the acceptance of smart bands in healthcare, this thesis builds on the

technology acceptance model and its variants TAM2 and UTAUT, as well as on trust

and technological personality that are important in this study context given the

specific issues of privacy, usability, and user characteristics of smart bands in

healthcare.

Thus, we propose a model similar to TAM, TAM2, and UTAUT, in which the actual

use (adoption) is influenced by behavioral intention to use (BI or acceptance) (Figure

5.1a). Similarly with TAM2 and UTAUT, the intention to use (BI) is directly

influenced by perceived usefulness (PU), perceived ease of use (PEOU), and social

influence (SI). In addition, we propose that trust and technological personality have

also direct influences on BI (Figure 5.1a). Table 5.1 defines all the constructs in the

model.

Figure 5.1a Research model of adoption of wearable devices in healthcare

The model can be extended to include three moderating variables: age, gender, and

behavioural motivation (Figure 5.1b). Age and gender are also present in UTAUT.

Behavioural motivation reflects personality and defines the inclination of people

towards withdrawing from or acting upon novel situations (see Gray, 1972). In

original UTAUT model, two more variables were included as moderating factors:

experience and voluntariness. Experience is defined in UTAUT similarly to mobile

24

technology skills in our multidimensional construct technological personality, and

thus we added the experience in another form as an independent factor. Voluntariness

means “the degree to which use of the innovation is perceived as being voluntary, or

of free will” (Moore and Benbasat, 1991). While social influence and innovativeness

have part meaning of voluntary and free will. Thus, we delete voluntariness from the

model.

Figure 5.1b Extended model of adoption of wearable devices in healthcar

25

Table 5.1. Definitions of key constructs in the proposed model.

Construct Definition References

Perceived Usefulness (PU) User believes that using wearable

devices would be beneficial to his or

her health.

adapted from

(Davis, 1989)

Perceived Ease of Use (PEOU) User believes that using wearable

devices would be free of effort.

adapted from

(Davis, 1989)

Trust User’s confidence in quality and

reliability of using wearable

devices.

adapted from

(Gefen et al., 2003)

Social influence Individual perceives that important

others believe he or she should use

wearable devices.

adapted from

(Venkatesh et al.,

2003)

Technological

personality

Mobile technology

skills

An individual judgement of one’s

capability to use mobile technology.

adapted from

(Compeau &

Higgins, 1995)

Innovativeness Individuals are willing to adopt

wearable devices or other related

applications that are totally new or

have the new functionality in the

context of their individual

experience.

(Aldaz et al. 2009)

Affinity The perceived importance of the

medium in the life of the individual

refers to wearable devices or its

applications.

(Rubin, 1981)

Compatibility The degree to which an innovation

is perceived as consistent with the

existing values, past experiences,

and needs of potential adopters

(Rogers, 1995).

Behavioural Intention to use (BI) User has formulated conscious plans

to use or not use wearable devices.

adapted from

(Davis, 1989)

Actual Use (AU) Behaviour use. adapted from

(Davis, 1989)

As shown in the preceding discussion, this study propose a hybrid technology

acceptance model to study users’ acceptance of the wearable devices in healthcare.

The proposed model consists of three categories of factors: 1) product intrinsic

features (PU, PEOU, and Trust), 2) social influence, and 3) user’s technological

personality. In this model, attitude toward using (ATT) is omitted (as it is also in

TAM2 and UTAUT) as its influence on behavioral intention to use and actual use is

weak (Venkatesh and Davis, 2000).

26

5.1 PU, PEOU, and Trust

A model of Trust and TAM was well defined in on-line shopping setting (Gefen et al.,

2003). Trust is one of the determinants of PU, PEOU is an antecedent of trust, and

trust has a direct influence on behavioral intention to use. Trust is defined as users’

confidence in quality and reliability. Thus, trust is supposed to influence the perceived

usefulness for the part of reliability, and perceived ease of use on behalf of quality of

wearable devices is supposed to influence users’ trust. The following hypotheses (Hs)

are derived based on the core of TAM, UTAUT, and trust models (H1 – H6):

H1. PU positively influences the behavioural intention to use wearable devices.

H2. PEOU positively influences the behavioural intention to use wearable devices.

H3. Trust positively influences the behavioural intention to use wearable devices.

H4. PEOU positively influences on PU to use wearable devices.

H5. PEOU positively influences trust on using wearable devices.

H6. Trust has positive effect on PU to use wearable devices.

5.2 Social influence

According to UTAUT, social influence affects the behavioral intention to use

technology. Social influence refers to individual perception that others expect that he

or she should use a technology, e.g., wearable devices. In UTAUT, social influence

affects BI through three mechanisms: compliance, internalization, and identification

(Venkatesh and Davis 2000; Warshaw 1980). Compliance usually happens inside an

organization; employees obey the rules to use technology offered. Thus, this construct

is not relevant to wearable devices such as smart bands in voluntary use contexts.

Internalization and identification refer to more voluntary uses of technology and are

triggered by other users’ recommendations, ads, former comments, and reputation. In

the case of wearable devices in health care, a doctor can recommend them, but still

the user is the one who decides. In our model, social influence refers to the meaning

of internalization and identification. Thus, social influence is expected to positively

influence acceptance of wearables devices (H7).

H7. Social influence positively influences on behavioural intention to use wearable

devices (BI).

27

5.3 Technological personality

Technological personality is defined as individual judgment of one’s capability to use

technology, one’s willing to adopt new technology, one’s feeling of how important of

technology is and what influence will technology bring (Compeau & Higgins, 1995;

Aldaz et al. 2009; Rubin, 1981; Rogers, 1995).

Under the context of technology, this paper adds technological personality including

mobile technology skills, innovativeness, affinity and compatibility.

Mobile technology skills

Koufaris (2002) highlighted the importance of web skills in modeling the behavioral

intention to use web technology, such as e-commerce. Similarly, mobile technology

skills affect the real interaction with mobile digital products. Thus, mobile technology

skills are expected to positively influence behavioral intention to use wearable devices

(H8).

H8. Mobile technology skills positively influences the behavioural intention to use

wearable devices.

In the model, mobile technology skills are defined as one’s capability to use mobile

technology by herself or himself (definition adapted from Compeau & Higgins,

1995).

Innovativeness

Innovativeness is possessed to a lesser or greater degree by all individuals (Varma,

Sprott, Silverman, & Stem, 2000). Considering that wearable devices are emerging

technologies appeared recently, this model adds innovativeness as a factor to evaluate

the acceptance of wearable devices (H9).

H9. Innovativeness positively influences the behavioural intention to use wearable

devices.

Affinity

Affinity was defined as the perceived importance of the medium in the life of the

individual (Rubin, 1981).

Aldaz-Manzano et al. (2009) showed that the effect of affinity is not strong in the

context of mobile shopping. However, in our model, affinity has been taken into

account given that wearable devices are emerging technologies and healthcare

applications are essential for people lives (see the Phillips survey, 2012). Thus, we

hypothesize that affinity influences the intention to use wearable devices (H10).

H10. Affinity positively influences the behavioural intention to use wearable devices.

28

Compatibility

Compatibility was defined as the degree to which an innovation is perceived as

consistent with the existing values, past experiences, and needs of potential adopters

(Rogers, 1995).

Aldaz-Manzano et al. (2009) showed that compatibility such as some prior experience

with an innovation or related fields has a positive effect on the acceptance of

emerging digital products, since the users may acquire concepts in their former

experience that push their present action for acceptance. M-shopping intention is

clearly influenced by individual perceived compatibility (Aldaz et al. 2009).

Analogizing the findings of the literature review, we hypothesize a similar influence

of compatibility on wearable devices acceptance (H11).

H11. Compatibility positively influences the behavioural intention to use wearable

devices.

5.4 Moderators

In UTAUT, there are four key moderating variables (experience, voluntariness, gender,

and age). The influence of performance expectancy, effort expectancy and social

influence on behavioural intention will be moderated by gender and age (Venkatesh et

al., 2003). In our extended model, we keep age and gender and they are expected to

moderate the following relationships: the influence of PU, PEOU, trust, social

influence and technological personality on behavioural intention to use (Figure 5.1b).

Our extended model omitted the voluntariness because social influence and

innovativeness have meaning of voluntariness. The moderating effect of experience

from effort expectancy, social influence and facilitating conditions on behavioural

intention is, in our model, mapped in an independent effect given by technological

personality construct, namely mobile technology skills. In addition, as earlier shown,

we add a new moderating variable, namely, behavioural motivation (behavioural

inhibition system/behavioural activation system, BIS/BAS).

Behavioural motivation

Behavioural motivation, reflected by the so-called behavioural inhibition system (BIS)

and behavioural activation system (BAS), underlies human behaviour. BIS and BAS

are conceptualized by Gray in early 1980s (1982, 1990) as two physiological

mechanisms that influence emotions and behaviour, and they also represent

personality traits; some people are more inclined toward withdrawal or inhibition

behaviour in certain circumstances, while other people are included towards approach

behaviour. The most common measures for BIS and BAS traits are the BIS/BAS

scales developed by Carver and White (1994). Each item of the BIS/BAS

questionnaire is a statement with options on a 4-point scale: 1 = very true for me; 2 =

somewhat true for me; 3 = somewhat false for me; 4 = very false for me. Carver and

29

White’s study shows that the BIS/BAs scale has high reliability and validity.

BAS is believed to regulate appetitive motives, and its goal is to move toward

something desired. BAS has three scales, Reward responsiveness (BAS RR), Drive

(BAS D), and Fun Seeking (BAS FS) that assess different aspects of BAS functioning.

BIS is said to regulate aversive motives, in which the goal is to move away from

something unpleasant. These scales have in total 20 items; 7 for BIS, 4 for BISD, 4

for BISFS, and 5 for BISRR. The BIS/BAS scales and the underlying constructs have

been used as moderating variables in studying approach motivation of users in

different contexts, such as reading from digital media (Rajanen et al. 2015), and

communication research (Kallinen et al., 2004; Ravaja, 2004). The results in these

studies indicated that people with higher BAS Drive and BAS FS experienced higher

approach motivation, that is, positive emotions. These, in turn, are believed to engage

people in actions and influence their behaviour. Thus, in our model, we included the

behavioural motivation as a moderating variable influencing the relationships between

BI and AU, as well between the independent factors and BI (see Figure 5.1b).

In this thesis, we focus on the model without moderators and the research hypotheses

are summarized in Figure 5.2.

Figure 5.2. Research model and hypothesis of adoption of wearable devices in healthcare

30

6. Research methodology

In this chapter, the quantitative research approach for this research is introduced. The

goals of this study are descriptive and explanatory research. For achieving the goals,

we employed quantitative research approach by developing a questionnaire and

conducting an online survey. A pilot study was conducted to refine the questionnaire.

This chapter also describes the methods of data collection and analysis.

6.1 Quantitative research approach

This study is a descriptive and explanatory research to find out users’ acceptance,

usage patterns, preferences with regard to product features, and the determinants of

users’ acceptance of wearable devices for healthcare. The research question is: what is

the state of acceptance of wearable technology for personal healthcare in China?

Quantitative research design was selected for answering the research question and for

testing research model of explaining acceptance of wearable devices for healthcare

(Creswell, 2009).

Descriptive research

This study is going to describe what is the state of acceptance of wearable technology

for personal healthcare in China? Through questionnaire, this study answers the

following sub-questions: what are the experiences of using smart bands and its

applications? What smart bands and what type of applications people usually use in

their daily lives? How long and how often for them to use smart bands?

Explanatory research

Another aim of the study is to test the research model of acceptance of wearable

devices in China and to find the relationships between independent and dependent

constructs (intention to use or acceptance and actual use or adoption). The

independent constructs are PEOU, PU, trust, social influence, and technological

personality. PU and trust are both independent and dependent (See Figure 5.2). A total

of 11 hypotheses are tested.

6.2 Questionnaire development

To collect data for answering the research questions for both the descriptive and

explanatory research goals, we developed a questionnaire based on the proposed

research model and literature reviewed (TAM, TAM2, UTAUT, trust model,

31

technological acceptance, and behavioural motivation).

The questionnaire has four parts: 1) technological personality questions, 2) usage

patterns (descriptive research), 3) research model constructs (explanatory research),

and 4) background information (including age, gender, behavioural motivation). The

full questionnaire is presented in Appendix.

Part 1 of the questionnaire explores also users’ experience with mobile phone,

including mobile technology skills evaluation adapted from Compeau and Higgins

(1995), innovativeness, affinity and compatibility adapted from the measurement

defined and modified by Aldaz-Manzano (2009).

Part 2 of questionnaire summarizes experience of smart bands and its applications that

only surveyed by the respondents who have used smart bands before.

Part 3 emphasizes the measurements in the research model, and contains 9 items.

Perceived usefulness and perceived ease of use are based on Davis’ technology

acceptance model (Davis, 1989). Trust is adapted from Gefen’s trust model (Gefen et

al., 2003). Social influence is based on Venkatesh’s UTAUT model (2003). The

construct of behavioural intention to use is adapted from Venkatesh and Davis (2000).

Part 4 of the questionnaire includes background information about the respondents,

and the BIS/BAS scales adapted from Carver and White (1994). In our

implementation, to keep the questionnaire short we only included one item per scale.

This questionnaire uses questions 7 (BAS RR), 9 (BAS D), 5 (BAS FS), and 16 (BIS)

according to their original item numbers, followed by the items.

BAS Reward Responsiveness (7): When I get something I want, I feel excited and

energized.

BAS Drive (9): When I want something I usually go all-out to get it.

BAS Fun Seeking (5): I'm always willing to try something new if I think it will be

fun.

BIS (16): If I think something unpleasant is going to happen I usually get pretty

"worked up."

Background information includes also age and gender.

Table 6.1 describes the items in Part 1 and Part 3 of the questionnaire. Each item is

evaluated on a 7-point Likert scale.

32

Table 6.1. Measurement items of model in survey

Construct Items Measurement items

Perceived usefulness PU1 Using the smart bands would help me monitor my physical

health.

PU2 I think using smart bands would help me improve my

physical health.

PU3 Using the smart bands would enhance my effectiveness in

monitoring my physical health.

PU4 Based on my perception of smart bands, I believe they

provide good features.

Perceived ease of use PEOU1 I think the interaction with the smart bands is clear and

understandable.

PEOU2 It would be easy for me to become skillful at using smart

bands.

PEOU3 I think it is easy to get the smart bands to do what I want it to

do.

PEOU4 I think that it takes low mental effort to use smart bands.

Trust TST1 I believe that the smart bands are reliable for the data

recording.

TST2 I believe that the personal information is safe.

TST3 I believe that the smart bands and their applications provide

accurate information.

Social influence SI1 My relatives and friends think that I should use smart bands.

SI2 Product ads influence me to use smart bands.

SI3 Former users' comments influence me to use smart bands.

Mobile technology skills MTS1 I am very skilled at using mobile technology.

MTS2 I know how to find what I want through mobile technology.

MTS3 I know more about using mobile technology than other users.

Innovativeness INO1 I think I know more about innovative digital products than

my circle of friends.

INO2 I think I would use an innovative digital product even if

nobody else I know had used it before.

Affinity AFF1 Using a mobile phone is one of my main daily activities.

AFF2 If my mobile phone is down, I really miss it.

AFF3 My mobile phone is important in my life.

AFF4 I would be lost without my mobile phone.

Compatibility COM1 How often do you search for health related information on

the Internet?

COM2 How often do you use health applications on your mobile

phone?

Behavioral intention to use

smart bands

BI1 I intend to use smart bands in the future.

BI2 I predict I would use smart bands in the future.

33

6.3 Pilot study

Because the questionnaire is going to research on the acceptance of smart bands in

China, the questionnaire will be filled by mostly people from China or living in China.

The questionnaire was developed in English and then translated into Chinese. For the

translation of the BIS/BAS scales, we used the Li, Zhang, Jiang, Li, Mi, Yi (2008)’s

Chinese version of the BIS/BAS scales for reference.

Before administering the questionnaire to the target population, a pilot study was

conducted to refine the questionnaire. Four Chinese who lived in Finland and China

were interviewed to test the Chinese version. The result of pilot study helped to

rephrase the sentences and refine the questions. For example, the question of whether

responders have mobile phone is changed to which type of mobile phone do they use

in daily life. Moreover, the interview for Chinese version led to slight modifications

of the selected BIS/BAS scales for the Chinese version; the final Chinese version was

done in the questionnaire website for data collection. The questionnaire of English

version showed in the Appendix is translated from final Chinese version.

6.4 Data collection

In this study, data are collected in a large sample questionnaire of smart bands’

adoption that experimented to empirically test this research model. The target

population in the study are people living in China and Chinese people living abroad.

Sampling method and questionnaire administration are introduced in this section.

6.4.1. Sampling

There are various kinds of sampling methods. For example, random sampling,

systematic sampling, stratified sampling, convenience sampling. Convenience

sampling (also known as availability or accidental sampling) is selected for this study.

Non-probability sampling techniques do not require random selection steps, just

involve a larger population of respondents that can be investigated (Tansey, 2007).

Convenience sampling is a kind of non-probability sampling. It is a statistical method

that people are selected just because ease of their volunteering or availability or easy

access. What’s more, researcher could have their convenience to select respondents

that are most readily available, no matter what characteristics they are or other settled

conditions, when the sample size reach the required sample size, the procedure of

sampling is done (Tansey, 2007). For the reason that questionnaire was posted in the

Internet and a social application, also researchers’ friends and relatives were invited.

Furthermore, author’s friends extend the questionnaire to their friends and social

circle. This type of sampling is known in the literature as snowball sampling (Ruane,

2005).

34

6.4.2. Questionnaire administration

The data for this study was gathered with a web-based questionnaire using the free

online tool Sojump. The link was posted on Weibo and Wechat, which are popular

social media in China, similar to Twiter and Line. The subject group for the

questionnaire was selected randomly in the social network site. The period for

opening questionnaire is a week between September 20, 2015 and September 27,

2015. After then, questionnaire and data are collected. The sample size was 158

respondents that chose to answer the survey.

6.5 Data analysis

In this section, the data analysis techniques are described, and the results are

presented in Chapter 7. For the descriptive research, namely for answering the

questions about usage patterns such as type of applications and frequency of use,

descriptive statistics are calculated in terms of mean values, frequency tables, and

graphs. For the explanatory research and hypothesis testing, structural equation

modeling is employed. Both structural model and measurement model are examined.

Analysis of structural model

In this research, we adopted Partial Least Squares Structural Equation Modeling

(PLS-SEM) for data analysis.

SEM (Structural Equation Modeling) is a statistical technique for testing and

estimating causal relationships among multiple independent and dependent constructs

based on statistical data and qualitative causal assumptions. (Gefen, Straub, &

Boudreau, 2000; Urbach & Ahlemann, 2010).

There are two types of SEM. Covariance-based SEM (CB-SEM) is primarily used to

confirm or reject theories. It does this by determining how well a proposed theoretical

model can estimate the covariance matrix for a sample data set. In contrast, Partial

Least Squares SEM (PLS-SEM) is a variance-based SEM method (also known as

component-based SEM), primarily used to develop theories in exploratory research. It

does this by focusing on explaining the variance in the dependent variables when

examining the model (Hair, Hult, Ringle, & Sarstedt, 2013). The goals of this study

are descriptive and explanatory research. For the descriptive part, descriptive statistics

are calculated; for the explanatory research, the factors which affecting behavioral

intention to use are explained. Thus, this study prefer PLS-SEM.

There are different software to assess Structural Equation Modeling, such as LISREL

and AMOS using Maximum Likelihood Estimate (MLE) for estimation of model

35

parameters. Krebsbach & Craig Michael (2014) point out that traditionally SEM

research involves large sample sizes of at least 200-400, or the SEM will instable

(Boomsma, 1983; Kenny, 2012; Kline, 2011). Wong (2013) based on prior research to

suggest that a sample size of 100 to 200 is usually a good starting point in carrying

out path modeling (Hoyle, 1995).

SmartPLS is one of the prominent software applications for PLS-SEM. It has friendly

user interface and advanced reporting features (Wong, 2013). By the reason of the

sample size is 158 (less than 200), we’d like to choose PLS-SEM to test the model

with the software of SmartPLS 2.0.

Analysis of measurement model

Validity usually behave as content validity, reliability, construct validity, manipulation

validity, and statistical conclusion validity (Straub, Boudreau & Gefen, 2004).

Content validity

As discussed in Straub (1989), content validity is established through literature

reviews and expert judges or panels. Empirical assessment of this validity is generally

not required.

Reliability

Internal Consistency usually measures through a variety of items in the same

instrumentation. Average Variance Extracted (AVE), Composite Reliability (CR) and

Cronach’s Alpha are usually the statistic most often used to evaluate internal

consistency (Straub et al., 2004). The AVE measures the percent of variance is

constructed, which reflect the proportion of the sum of the variance captured by the

construct and measurement variance (Gefen, Straub, & Boudreau, 2000). Usually the

AVE value has to exceed the generally recognized standard: 0.50 (Fornell & Larcker,

1981).

The generally composite reliability value is better to be 0.70 or greater (Fornell &

Larcker, 1981), while the recommended value for a Cronbach’s alpha is also 0.70 or

greater (Gefen et al. 2000).

Construct validity

As pointed out by Straub (1989), construct validity differs from internal validity that

focuses on the correspondence between measurements and the concepts or constructs

while internal validity focuses on the extent to which a causal conclusion could be

attributed to the treatment that could define methodological quality.

36

7. Results

In this chapter, results are analyzed through sample characteristics, analysis of

descriptive research and exploratory research. And also the results will be interpreted

with both measurement model and structural model detailed. The PLS analysis assess

both the measurement model and structural model based on the data collected from

158 responses.

7.1 Sample characteristics

The sample characteristics are shown in Table 7.1. As we can see from the sample

demographic, respondents have the gender balance, so the data are reliable. Most of

the respondents are young people in the age range 18 to 29. There are two respondents

from other countries, we include them in the data for the reason that they are Chinese

living in Finland. 92.4% of the respondents have college, graduate school or higher

education.

Table 7.1. Sample characteristics

Variable Category Frequency Percent (%)

Gender Male 68 43.04

Female 90 56.96

Country of residence China 156 98.73

Other country 2 1.27

Age

18-24 89 56.33

25-29 40 25.32

30-34 9 5.7

35-39 6 3.8

40-44 2 1.27

45-49 6 3.8

>50 6 3.8

Education

Under high school 4 2.53

High school 8 5.06

College 100 63.29

Graduate school (or above) 46 29.11

Monthly income

¥ 0 – 4000 75 47.47

¥ 4001 – 8000 50 31.65

¥ 8001 – 12000 20 12.66

¥ 12001 + 13 8.23

37

Table 7.2 describes the respondents’ fields of work; 32.91% of respondents work in

IT.

Table 7.2. Fields of work


Field of Work

IT 52 32.91%

Finance, bank, economics and consulting 18 11.39%

Education 7 4.43%

Industry 6 3.80%

Marketing, advertisement and trade 4 2.53%

Health 4 2.53%

Government 4 2.53%

Architecture, Engineering and Design 4 2.53%

Sports 3 1.90%

Secretary 2 1.27%

Media 2 1.27%

Tourism 1 0.63%

Networks 1 0.63%

Law 1 0.63%

Entrepreneur 1 0.63%

Other 48 30.38%

7.2 Smart bands usage patterns

Table 7.3 shows the usage patterns of smartphones and smart bands. 96.2% of

respondents use smartphone in daily life. 35 respondents (22.16%) have experience of

using smart bands (16 use smart bands currently, and 19 abandoned). 70 respondents

(44.3%) have not used, but are interested.

Table 7.3. Experience with mobile phone and smart bands (n= 158).


Type of mobile phone Basic mobile phone 6 3.8

Smartphone 152 96.2

Experience of using smart

bands

Yes and still using now. 16 10.13

Yes, but abandoned. 19 12.03

No, but interested to. 70 44.3

No, never wanted to use. 53 33.54

Table 7.4 shows the usage patterns of the 35 respondents who have experience with

smart bands. As to the type of applications used frequently, sport activities accounts

for 88.57% while sleep monitoring accounts for 71.43%. What's more, 22 of 35

usually use or used smart bands daily. The most popular smart bands are Mi (10) and

Apple (8); the third most preferred was Fitbit (5).

38

Table 7.4. Usage patterns of smart bands (n= 35).


Most frequently used

smart bands

Jawbone 1 2.86

Fitbit 5 14.29

Misfit 1 2.86

Nike 1 2.86

Mi 10 28.6

Apple 8 22.86

Polar 1 2.86

Coolpad 1 2.86

HUAWEI 2 5.71

BONG 2 5.71

Others 3 8.57

Type of applications used

frequently

Sport activities 31 88.57

Sleep monitoring 25 71.43

Heart rate recording 10 28.57

Others 4 11.43

How long time use a smart

band

Less than a week 1 2.86

A few weeks, but less than a month 5 14.29

1-2 months 5 14.29

3-5 months 7 20

6-11 months 13 37.14

12-24 months 4 11.43

How often use a smart

band

Daily 22 62.86

Several times a week 8 22.86

Every few weeks 2 5.71

Once in few months or more rarely 3 8.57

Figure 7.1 shows that the percentages of using smart bands frequently are nearly the

same regardless the type of applications. There is no evidence support to that users’ usage patterns (type of applications) influence past users to abandon the use of smart

bands.

Figure 7.1. Type of smart bands related applications used frequently (%)

0.94

0.75

0.31

0.13

0.84

0.68

0.26

0.11

0.00

0.20

0.40

0.60

0.80

1.00

Sport activities Sleep monitoring Heart raterecording

Others

current users (n=16) past users (n=19)

39

Figure 7.2 shows that most of past users have used smart bands for 3 to 5 months

while most of current users use for 6 to 11 months.

Figure 7.2. How long time use smart bands

Figure 7.3 presents that the percentage of current users (75%) who use smart bands

daily are more than past users (53%), and the rest current users prefer several times a

week.

Figure 7.3. How often use smart bands (%)

Table 7.5 presents the results of BIS/BAS scales BIS/BAS (1 very true for me – 4

very false for me). It shows that in the sample, there is a tendency towards BAS

Reward responsiveness.

Table 7.5. BIS/BAS profiles of respondents

Variable Confidence Interval of Average Mean SD

BIS 2.188 to 2.470 2.33 0.899

BAS FS 1.888 to 2.175 2.03 0.913

BAS RR 1.744 to 2.066 1.91 1.03

BAS D 2.027 to 2.302 2.16 0.873

1

4

2 1

7

1 0

1

3

6

4 3

012345678

Less than aweek

A few weeks,but less than

a month

1-2 months 3-5 months 6-11 months 12-24 months


0.75

0.25 0.00 0.00

0.53

0.21 0.11 0.16 0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

Daily Several times aweek

Every few weeks Once in few monthsor more rarely


40

Table 7.6 presents correlation coefficient (Pearson’s) between BIS/BAS scales and BI.

The results show that there are no significant correlations between BIS/BAS items

and BI.

Table 7.6. Correlation coefficient between BIS/BAS scales and BI

Correlations coefficient between BIS/BAS scales and BI

BI BIS BAFS BARR BASD

BI Pearson Correlation 1 -.066 .019 .003 .042

Sig. (2-tailed) .409 .817 .972 .603

N 158 158 158 158 158

**. Correlation is significant at the 0.01 level (2-tailed).

7.3 Evaluation of measurement model

Validity usually is defined as content validity, reliability and construct validity

Content validity

In this research, literature reviews have been done to inform the development of the

measurement model and constructs. All the constructs and items in the model are

adapted form reliable papers and established models (TAM, TAM2, UTAUT,

technological personality, and trust).

Reliability

The average variance extracted (AVE) value has to exceed the generally recognized

standard: 0.50 (Fornell & Larcker, 1981). The value of composite reliability (CR) has

to be equal or greater than 0.70 (Fornell & Larcker, 1981), while the recommended

value for a Cronbach’s alpha should be 0.70 or greater (Gefen et al. 2000).

In this study, the AVE values are greater than 0.64, CR values range from 0.84 (for

social influence) to 0.94 (for mobile technology skills and compatibility), and

Cronbach’s Alphas are from 0.72 (for social influence) to 0.90 (for mobile technology

skills), all exceeding the recommended value (Table 7.7). Thus, the measurement

model has good internal consistency and reliability.

41

Table 7.7 Average Variance Extracted (AVE), Composite Reliability (CR) and Cronbach’s

Alpha of the measurement model

Construct Number of items AVE CR Cronbach’s

Alpha

PU 4 0.76 0.93 0.89

PEOU 4 0.72 0.91 0.87

TST 3 0.70 0.88 0.79

SI 3 0.64 0.84 0.72

MTS 3 0.84 0.94 0.90

INO 2 0.85 0.92 0.83

AFF 4 0.75 0.92 0.89

COM 2 0.89 0.94 0.88

BI 2 0.80 0.89 0.75

Construct validity

Researchers usually use MTMM (multitrait-multimethod matrices) to evaluate and

discriminant validity. The figures of diagonal line mean the square root of AVEs

(indicated in bold underlined) and the table shows the comparison between the square

root of AVEs and latent variable correlation. The figures that are not in diagonal line

represent correlation coefficients between one variable and others. Table 7.8 shows

that for all the constructs, the square root of AVE is greater than the construct’s

correlations with other constructs, demonstrating discriminant validity, thus construct

validity.

Table 7.8. The square root of AVEs, latent variable correlation of the measurement model

PU PEOU TST SI MTS INO AFF COM BI

PU 0.87

PEOU 0.65 0.85

TST 0.67 0.63 0.84

SI 0.61 0.55 0.54 0.80

MTS 0.43 0.62 0.36 0.29 0.92

INO 0.42 0.54 0.48 0.38 0.66 0.92

AFF 0.43 0.47 0.45 0.30 0.58 0.44 0.87

COM 0.29 0.37 0.43 0.33 0.24 0.21 0.25 0.94

BI 0.64 0.53 0.53 0.61 0.38 0.41 0.42 0.39 0.89

42

Table 7.9. Factor loadings and cross-loadings of research model

PU PEOU TST SI MTS INO AFF COM BI

PU1 0.91 0.58 0.62 0.55 0.40 0.48 0.40 0.21 0.62

PU2 0.84 0.51 0.57 0.54 0.32 0.28 0.35 0.19 0.49

PU3 0.90 0.60 0.57 0.57 0.39 0.37 0.41 0.27 0.55

PU4 0.83 0.57 0.56 0.45 0.39 0.33 0.31 0.32 0.56

PEOU1 0.63 0.87 0.53 0.53 0.56 0.42 0.42 0.34 0.42

PEOU2 0.52 0.89 0.53 0.54 0.58 0.53 0.42 0.32 0.50

PEOU3 0.54 0.81 0.56 0.36 0.46 0.41 0.36 0.29 0.42

PEOU4 0.50 0.82 0.53 0.42 0.48 0.46 0.39 0.32 0.46

TST1 0.62 0.57 0.87 0.45 0.39 0.48 0.41 0.40 0.52

TST2 0.54 0.49 0.80 0.49 0.27 0.36 0.36 0.38 0.45

TST3 0.49 0.53 0.84 0.40 0.24 0.35 0.37 0.30 0.35

SI1 0.52 0.43 0.45 0.79 0.23 0.31 0.19 0.24 0.48

SI2 0.46 0.36 0.39 0.84 0.13 0.24 0.16 0.35 0.50

SI3 0.48 0.53 0.45 0.77 0.33 0.35 0.36 0.20 0.49

MTS1 0.42 0.61 0.34 0.28 0.93 0.60 0.58 0.22 0.37

MTS2 0.36 0.56 0.31 0.18 0.89 0.49 0.58 0.25 0.30

MTS3 0.41 0.53 0.34 0.31 0.92 0.69 0.45 0.19 0.38

INO1 0.39 0.53 0.41 0.31 0.67 0.92 0.42 0.23 0.36

INO2 0.39 0.47 0.47 0.38 0.55 0.93 0.39 0.16 0.40

AFF1 0.39 0.41 0.35 0.20 0.58 0.40 0.86 0.21 0.39

AFF2 0.41 0.43 0.46 0.30 0.50 0.40 0.92 0.25 0.40

AFF3 0.36 0.43 0.37 0.24 0.57 0.40 0.93 0.23 0.37

AFF4 0.32 0.35 0.39 0.31 0.33 0.31 0.74 0.17 0.30

COM1 0.23 0.32 0.39 0.26 0.27 0.23 0.20 0.93 0.32

COM2 0.30 0.38 0.43 0.35 0.19 0.18 0.26 0.96 0.40

BI1 0.64 0.45 0.57 0.55 0.35 0.40 0.41 0.34 0.91

BI2 0.49 0.50 0.37 0.54 0.34 0.33 0.34 0.36 0.88

Table 7.9 shows the factor loadings ((indicated in bold) for the items are not only

greater than 0.7, but higher than cross-loadings which means that the more variance is

shared between the measures their own construct than with other constructs. Thus, the

internal consistency and discriminant validity are both acceptable.

7.4 Evaluation of structural model

In this section, path coefficients of the structural model are estimated using smartPLS

2.0. Hypotheses H1 - H11 are tested through bootstrapping and replicated sample

method with sample 5000. The significance levels of path coefficients are also tested.

43

Figure 7.1 and table 7.10 show the results. Figure 7.1 shows that R square of

perceived usefulness, trust and behavioral intention are 0.53, 0.40 and 0.54. Thus, the

model overall fitted quite well.

Figure 7.4. PLS result of research model

PU, PEOU, and Trust

Hypothesis testing shows the effect from perceived usefulness on behavioral intention

are significant (β = 0.34, p < 0.001); The effect from perceived ease of use on

perceived usefulness are significant (β = 0.65, p < 0.001); The effect from trust on

perceived usefulness are significant (β = 0.43, p < 0.001); The effect from perceived

ease of use on trust are significant (β = 0.63, p < 0.001). Thus, H1, H4, H5, H6 are

supported.

As can be seen from Figure 7.4 about TAM and trust model, the relationship between

perceived ease of use, trust and behavioral intention is not significant. Nevertheless,

the effect from perceived ease of use, trust on perceived usefulness and the effect

from perceived ease of use on trust are extremely significant (p < 0.001).

However, the relationship between perceived ease of use and behavioral intention is

not significant, H2 is not supported. The relationship between trust and behavioral

intention is not significant, H3 is not supported. Thus, we consider the influence on

behavioral intention from trust and perceived ease of use are not directly. Trust and

perceived ease of use affect behavioral intention indirectly from perceived usefulness.

44

Social Influence

The effect of social influence on behavioral intention is significant (β = 0.30, p <

0.01), supporting H7.

Technological personality

The effect of affinity on behavioral intention (β = 0.13, p < 0.05), and the effect of

compatibility on behavioral intention are significant (β = 0.15, p < 0.005), thus H10

and H11are supported.

The relationship between mobile technology skills and behavioral intention is not

significant; H8 is not supported. The relationship between innovativeness and

behavioral intention is not significant, H9 is not supported.

Figure 7.5 presents that there is no significant linear correlation between technological

personality and behavioral intention to use smart bands.

Figure 7.5. Scatter plots (technological personality and BI)

Reasons of this phenomenon and limitations will be discussed in next chapter.

45

Table 7.10. Result of hypothesis testing

Hypothesis path Path

Coefficients

Standard

Deviation T Statistics

Significance

level Result

H1 Perceived usefulness Behavioral intention 0.34 0.09 3.68 p < 0.001 Supported

H2 Perceived ease of use Behavioral intention 0.21 0.12 0.01 p >0.05 Not supported

H3 Trust Behavioral intention 0.13 0.10 0.10 p >0.05 Not supported

H4 Perceived ease of use Perceived usefulness 0.65 0.09 3.98 p < 0.001 Supported

H5 Perceived ease of use Trust 0.63 0.06 11.25 p < 0.001 Supported

H6 Trust Perceived usefulness 0.43 0.08 5.15 p < 0.001 Supported

H7 Social influence Behavioral intention 0.30 0.09 3.22 p < 0.01 Supported

H8 Mobile technology skills Behavioral intention -0.01 0.09 0.08 p >0.05 Not supported

H9 Innovativeness Behavioral intention 0.08 0.08 0.90 p >0.05 Not supported

H10 Affinity Behavioral intention 0.13 0.06 2.10 p < 0.05 Supported

H11 Compatibility Behavioral intention 0.15 0.07 2.02 p < 0.05 Supported

46

8. Discussion

In this chapter the results of the study will be discussed. First implications will be

discussed when compared to other studies. In addition, based on the research results,

suggestions will be presented according to researcher, professionals, developers of

smart bands, and also marketing strategy mangers. Next the limitations of the research

will be evaluated. Finally some suggestions for future work will be presented in the

end of this chapter.

8.1 Implications of results

This study investigated technology acceptance of wearable devices focused on smart

bands. The goal of the study was to provide a state-of-the-art of the adoption and

acceptance of smart bands for personal healthcare in China. The main research

question of this study is: What is the state of acceptance of wearable technology for

personal healthcare in China?

Based on the previous research, 11 hypothesizes were formulated. In addition, a

research model was formulated based on these hypothesizes that were tested and

analyzed (Figure 5.2). And the results showed in Figure 7.4 have the conclusion that

factors affecting user’ behavioral intention to use smart bands directly are: perceived

usefulness, social influence, affinity, and compatibility. As discussed in Chapter 7,

trust and perceived ease of use affect behavioral intention indirectly from perceived

usefulness. Another reason why effect on behavioral intention from perceived ease of

use is not supported is that, with the development of mobile technology, people are

getting familiar with emerging technologies. Thus, the influence from perceived ease

of use become weak. Not as expected, the influence from mobile technology skills

and innovativeness are not significant, there is no significant linear correlation

between technological personality and behavioral intention to use smart bands (see

Figure 7.5).

Jere, Koivumäki and Lappi (2015) studied consumers’ acceptance of future My Data

based preventive eHealth services with adapted model from UTAUT2. In their study,

factors affecting behavioral intention are performance expectancy, effort expectancy,

social influence, facilitating conditions, habit, vulnerability, and self-efficacy. They

also discussed in their study that the effect on behavioral intention from effort

expectancy was weak because most respondents are particularly familiar with the

Internet technology. Similarly, the majority of respondents are quite young and related

with IT in our research.

47

Lu, Liu, Yu and Wang (2008) investigated the factors influencing adoption of wireless

mobile data services (WMDS) in China based on TAM and related constructs

including trust, WMDS technology, personal innovativeness in information

technology (PIIT), and mobile trust belief. Their findings showed that WMDS

adoption intention is determined by consumers’ perceived usefulness and perceived

ease of use of WMDS mainly. What’s more, it also relied on personal innovativeness

in information technology and mobile trust belief. Compare to our research,

innovativeness and trust’s influence on behavioural intention to use smart bands are

not supported.

For the structure of the research model, compare to other studies, this study have link

TAM2 and UTAUT with trust model, and also related to the technological personality

(mobile technology skills, innovativeness, affinity and compatibility) to create a

combined model. As in TAM2 and UTAUT, the model omitted attitude toward using,

as its influence on behavioral intention to use and actual use is weak (Venkatesh and

Davis, 2000).

Usually in UTAUT, there are four key moderating variables (experience,

voluntariness, gender, and age). This study remove voluntariness in the model, and

keep gender, age and add BIS/BAS scales as three moderators. Experience is

contained inside technological personality that removed from moderators. This

remains moderating effects testing for future work.

For researchers, in the future, the research model can be replicated with larger

sampling size and other sampling method. For healthcare professionals, smart bands

can also be replaced with other wearable devices, and this model can be applied to

exploring other wearable devices. For developers, users mainly will be influenced by

perceived usefulness, they need to focus on the functionality and features.

Regarding the usage patterns, this study showed that users abandoned the smart bands

after using the technology for few weeks up to 12-24 months (See Figure 7.2). Most

of the past users abandoned the smart bands after 3-5 months of usage (31.5% within

3-5 months of usage; 21% within 6-11 months; and 15.8% within 12-24 months). The

percentage of users abandoning the smart bands within a month is much less in this

study compared to the study by Tencent in 2014 (5% versus 45.7%). This shows a

trend in popularity of smart bands, but also that smart bands are still abandoned even

after a long period of use. For future research is important to study the reasons of

abandoning the smart bands.

8.2 Limitations of studies and future work

It cannot be denied that this study has some limitations. One limitation is the

convenient sampling method for data collection; According to Robson (1993),

48

whoever gets sampled is determined by all kinds of unspecifiable biases and

influences (Farrokhi, & Mahmoudi-Hamidabad, 2012). Future work could consider

more ways of distributing the link of questionnaire, and try to involve more volunteers

to participate with longer duration for opening questionnaire. Furthermore, only 35 of

158 respondents have the experience of using smart bands is another limitation; but

on the other hand this can be the true extent of adoption of smart bands at the current

moment in September 2015. Due to the convenience sampling, the sample consist of

mainly young people working in the IT field. Thus for generalizing the results, is

needed to replicate the study on a larger and more heterogeneous sample.

In the future, the relationship between BIS/BAS scales could be analyzed in detail to

test the moderating effects of personality on the relationships in the research model.

Moreover, the respondents can be selected according to different requirements (age,

gender, mobile technology skills and experience). Based on the review on user

characteristics, there are more constructs can be added in the technological personality,

such as attitudes towards privacy related concerns. For instance, privacy can be an

antecedent of trust while privacy can be a determinant of trust. What’s more,

definition of usability on wearable devices (or smart bands) still not clear and definite,

although PU and PEOU contains most of the contents. In addition, future research

should explore with qualitative methods to investigate which functionality can be

mainly include in the wearable devices that can be accepted by users.

49

9. Conclusion

The aim of this study is to examine the acceptance of wearable devices focused on

smart bands, in order to find the state of acceptance of wearable technology for

personal healthcare in China. Although technology acceptance model has been studied

when combined to trust model and technological personality related, separately, this

study try to make a contribution to combine TAM with trust model and technology

personality based on TAM 2 and UTAUT, and accompanied with the moderators of

age, gender, and behavioral motivation.

This study is a descriptive and explanatory research that describe usage patterns and

the extent of adoption and acceptance of wearable devices in China, and explain

which factors influence behavioral intention to use wearable devices for healthcare.

Based on previous research, literature reviewed on wearable devices, especially in the

fields of healthcare, and draw the concepts related to user evaluation. Then this study

mainly conducted technology acceptance model related theories and technological

personality to put forward a research model with 11 hypothesis. The research model

had been tested through an online questionnaire, and 158 completed responses were

received. Using Partial Least Squares Structural Equation Modelling (PLS - SEM),

research model was tested in smartPLS software.

Findings showed that 96.2% of respondents use smartphone in daily life, and 44.3%

of respondents have not used, but are interested in using smart bands while 12.03%

abandoned to use. Sport activities accounts for 88.57% and sleep monitoring accounts

for 71.43% when asked which type of applications use frequently.

The factors affecting user’ behavioral intention to using smart bands directly are:

perceived usefulness, social influence, affinity, and compatibility. Trust and perceived

ease of use affect behavioral intention indirectly through perceived usefulness.

However, contrary to expectations, the influence from mobile technology skills and

innovativeness are not significant.

The study is significant because it provides information on usage patterns of the

emerging smart bands for personal healthcare. Moreover, the determining factors of

acceptance among potential customers can give developers a feedback for the future

improvement of wearable devices. Based on the research of users’ adoption towards

wearable devices in healthcare can help related companies have its needs-based

positioning and make special marking strategies to reach their marketing target.

50

Reference

Afthanorhan, W. M. A. B. W. (2013). A comparison of partial least square structural

equation modeling (PLS-SEM) and covariance based structural equation

modeling (CB-SEM) for confirmatory factor analysis. International Journal

Engineering and Science Innovative Technologies (IJESIT), 2(5), 8.

Ajzen, I., & Fishbein, M. (1980). Understanding Attitudes and Predicting Social

Behaviour. Englewood Cliffs, NJ: Prentice-Hall.

Aldás-Manzano, J., Ruiz-Mafe, C., & Sanz-Blas, S. (2009). Exploring individual

personality factors as drivers of M-shopping acceptance. Industrial Management

& Data Systems, 109(6), 739-757.

Anderson, J., & Rainie, L. (2014). The Internet of things will thrive by 2025. Pew

Research Internet Project, 14.

Andrew Michael. (2013). Sign Language Ring Translates Hand Movements Into

Spoken Words. Wearable Devices.com, December 5, 2013. Retrieved March 30,

2016 from

http://www.wearabledevices.com/2013/12/05/sign-language-ring-translates-hand

-movements-spoken-words/

Ann Carrns. (2013). Free Apps for Nearly Every Health Problem, but What about

Privacy? N.Y. TIMES. Retrieved March 30, 2016 from

http://www.nytimes.com/2013/09/12/your-money/free-apps-for-nearlyevery-heal

th-problem-but-what-about-privacy.html

Billinghurst, M., & Starner, T. (1999). Wearable devices: new ways to manage

information. Computer, 32(1), 57-64.

Boomsma, A. (1983). On the robustness of LISREL (maximum likelihood estimation)

against small sample size and no normality. Phd thesis, University of Groningen.

Carver, C. S., & White, T. L. (1994). Behavioral inhibition, behavioral activation, and

affective responses to impending reward and punishment: The BIS/BAS scales.

Journal of Personality and Social Psychology, 67, 319-333.

http://www.nytimes.com/2013/09/12/your-money/free-apps-for-nearlyevery-health-problem-but-what-about-privacy.html

http://www.nytimes.com/2013/09/12/your-money/free-apps-for-nearlyevery-health-problem-but-what-about-privacy.html

51

Chau, P. Y., & Hu, P. J. (2002). Examining a model of information technology

acceptance by individual professionals: An exploratory study. Journal of

management information systems, 18(4), 191-229.

Compeau, D. R., & Higgins, C. A. (1995). Computer self-efficacy: Development of a

measure and initial test. MIS quarterly, 189-211.

Creswell, J. (2009). Research design: Qualitative, quantitative, and mixed methods

approaches. SAGE Publications, Incorporated.

Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance

of information technology. MIS quarterly, 319-340.

Deng, Z., Mo, X., & Liu, S. (2014). Comparison of the middle-aged and older users’

adoption of mobile health services in China. International journal of medical

informatics, 83(3), 210-224.

Eddy, N. (2014). Smart Devices, Wearables Pose Security Risks for Consumers.

Eweek, 6.

Endomondo. (2015). Endomondo features. Retrieved August 9, 2015 from

https://www.endomondo.com/features

Farrokhi, F., & Mahmoudi-Hamidabad, A. (2012). Rethinking convenience sampling:

Defining quality criteria. Theory and practice in language studies, 2(4), 784.

Fitbit, the NPD Group. (2015). Monthly Retail Tracking Service: Jan-Mar, Digital

Fitness Devices data - GPS enabled watches. Retrieved August 9, 2015 from:

https://www.fitbit.com/

Flaherty, J. L. (2014). Digital diagnosis: privacy and the regulation of mobile phone

health applications. Am. JL & Med., 40, 416.

Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with

unobservable variables and measurement error. Journal of marketing research,

39-50.

Gaul, S., & Ziefle, M. (2009). Smart home technologies: Insights into

generation-specific acceptance motives (pp. 312-332). Springer Berlin

Heidelberg.

Gefen, D. (2004). What makes an ERP implementation relationship worthwhile:

Linking trust mechanisms and ERP usefulness. Journal of Management

Information Systems, 21(1), 263-288.

https://www.endomondo.com/features

https://www.fitbit.com/

52

Gefen, D., Karahanna, E., & Straub, D. W. (2003). Trust and TAM in online shopping:

an integrated model. MIS quarterly, 27(1), 51-90.

Gefen, D., Straub, D., & Boudreau, M. C. (2000). Structural equation modeling and

regression: Guidelines for research practice. Communications of the association

for information systems, 4(1), 7.

Gray, J. A., & McNaughton, N. (2003). The neuropsychology of anxiety: An enquiry

into the function of the septo-hippocampal system (No. 33). Oxford university

press.

Gray, J. A. (1990). Brain systems that mediate both emotion and cognition. Cognition

& Emotion, 4(3), 269—288.

Hair Jr, J. F., Hult, G. T. M., Ringle, C., & Sarstedt, M. (2013). A primer on partial

least squares structural equation modeling (PLS-SEM). Sage Publications.

Hemmadi, M. (2014). How workplace wearables will make your employees more

productive. Canadian Business, 87(13), 12-14.

Holden, R. J., & Karsh, B. T. (2010). The technology acceptance model: its past and

its future in health care. Journal of biomedical informatics, 43(1), 159-172.

Hoyle, R. H. (Ed.). (1995). Structural equation modeling: Concepts, issues, and

applications. Sage Publications.

Hung, K., Zhang, Y. T., & Tai, B. (2004, September). Wearable medical devices for

tele-home healthcare. In Engineering in Medicine and Biology Society, 2004.

IEMBS'04. 26th Annual International Conference of the IEEE (Vol. 2, pp.

5384-5387). IEEE.

Jakob Nielsen. (2012). Usability 101: Introduction to Usability. Nielsen Norman

Group. Retrieved March 30, 2016 from

http://www.nngroup.com/articles/usability-101-introduction-to-usability/

Jawbone. (2015). Compare trackers. Retrieved August 9, 2015 from

https://jawbone.com/up/trackers

Jere, V., Koivumäki, T., & Lappi, M. (2015). Consumer acceptance of future My Data

based preventive eHealth services.

http://www.nngroup.com/articles/usability-101-introduction-to-usability/

https://jawbone.com/up/trackers

53

Jung, C., & Padman, R. (2014). Virtualized healthcare delivery: Understanding users

and their usage patterns of online medical consultations. International journal of

medical informatics, 83(12), 901-914.

Kenny, D. A. (2012). Measuring model fit. Retrieved electronically October 22, 2012

from

http://www.davidakenny.net/cm/fit.htm

Kallinen, K., Ravaja, N. (2004). Emotion-related effects of speech rate and rising vs.

falling background music melody during audio news: The moderating influence

of personality. Personality and individual differences, 37(2), 275--288.

Kharrazi, H., Chisholm, R., VanNasdale, D., & Thompson, B. (2012). Mobile

personal health records: an evaluation of features and functionality. International

journal of medical informatics, 81(9), 579-593.

Kline, R. B. (2015). Principles and practice of structural equation modeling. Guilford

publications.

Knight, J. F., Deen-Williams, D., Arvanitis, T. N., Baber, C., Sotiriou, S.,

Anastopoulou, S., & Gargalakos, M. (2006, October). Assessing the wearability

of wearable computers. In Wearable Computers, 2006 10th IEEE International

Symposium on (pp. 75-82). IEEE.

Knoblauch, M. (2014). The History of Wearable Tech, From the Casino to the

Consumer. Mashable. com, May, 13.

Koufaris, M. (2002). Applying the technology acceptance model and flow theory to

online consumer behavior. Information systems research, 13(2), 205-223.

Krebsbach, Craig Michael. (2014). Bootstrapping with small samples in structural

equation modeling: goodness of fit and confidence intervals. Open Access

Master's Theses. Paper 165.

Li, J., Talaei-Khoei, A., Seale, H., Ray, P., & MacIntyre, C. R. (2013). Health care

provider adoption of e- Health: systematic literature review. Interactive Journal

of Medical Research, 2(1).

Li, Y. Z., Zhang, Y., Jiang, Y., Li, H., Mi, S., & Yi, G. J. (2008). The Chinese version

of the BIS/BAS scale: reliability and validity. Chinese Mental Health Journal,

22(8), 613-616.

Lu, J., Liu, C., Yu, C. S., & Wang, K. (2008). Determinants of accepting wireless

mobile data services in China. Information & Management, 45(1), 52-64.

http://www.davidakenny.net/cm/fit.htm

54

Lymberis, A. (2003, September). Smart wearable systems for personalised health

management: current R&D and future challenges. In Engineering in Medicine

and Biology Society, 2003. Proceedings of the 25th Annual International

Conference of the IEEE (Vol. 4, pp. 3716-3719). IEEE.

Mackert, M., Champlin, S. E., Holton, A., Muñoz, I. I., & Damásio, M. J. (2014).

eHealth and health literacy: A research methodology review. Journal of Computer‐Mediated Communication, 19(3), 516-528.

Magnezi, R., Grosberg, D., Novikov, I., Ziv, A., Shani, M., & Freedman, L. S. (2015).

Characteristics of patients seeking health information online via social health

networks versus general Internet sites: a comparative study. Informatics for

Health and Social Care, 40(2), 125-138.

Magno, M., Porcarelli, D., Brunelli, D., & Benini, L. (2014, November). InfiniTime:

A multi-sensor energy neutral wearable bracelet. In Green Computing

Conference (IGCC), 2014 International (pp. 1-8). IEEE.

Mann, S. (2001). Can humans being clerks make clerks be human? Information

stechnik und Technische Informatik, 43(2), 97-106.

Marcengo, A., & Rapp, A. (2014). Visualization of human behavior data: the

quantified self. Innovative approaches of data visualization and visual

analytics, 1, 236-265.

miCoach. (2015). Retrieved August 9, 2015 from

http://micoach.adidas.com/

Misfit Wearables. (2015). Specifications. Retrieved August 9, 2015 from

http://misfit.com/products/

Moore, G. C., & Benbasat, I. (1996). Integrating diffusion of innovations and theory

of reasoned action models to predict utilization of information technology by

end-users. In Diffusion and adoption of information technology (pp. 132-146).

Springer US.

Nike+ running. (2015). Retrieved August 9, 2015 from

https://secure-nikeplus.nike.com/plus/

Parsons T. (2009). Johns Hopkins Bloomberg School of Public Health. Retrieved

August 24, 2015, from

http://www.jhsph.edu/news/news-releases/2009/gibbons-consumer-health.html

http://micoach.adidas.com/

http://misfit.com/products/

https://secure-nikeplus.nike.com/plus/

http://www.jhsph.edu/news/news-releases/2009/gibbons-consumer-health.html

55

Pavlou, P. A. (2003). Consumer acceptance of electronic commerce: Integrating trust

and risk with the technology acceptance model. International journal of

electronic commerce, 7(3), 101-134.

Punagin, S., & Arya, A. (2015). Privacy in the age of Pervasive Internet and Big Data

Analytics–Challenges and Opportunities. International Journal of Modern

Education and Computer Science (IJMECS), 7(7), 36.

PwC. (2014). The Wearable Future. Retrieved March 30, 2016, from

https://www.pwc.com/mx/es/industrias/archivo/2014-11-pwc-the-wearable-future

.pdf.

Rajanen, D., Salminen, M., & Ravaja, N. (2016, January). Reading a Newspaper on

Print Versus Screen: A Motivational Perspective. In 2016 49th Hawaii

International Conference on System Sciences (HICSS) (pp. 630-637). IEEE.

Ravaja, N. (2004). Effects of a small talking facial image on autonomic activity: The

moderating influence of dispositional BIS and BAS sensitivities and emotions.

Biological Psychology, 65(2), 163—183.

Robert Morris. (2015). Wearable Technology. Function, Fit, Fashion, Fun. Retrieved

29.3.2016 from

http://www.onebeacontech.com/OneBeaconTech/pages/news

Robson, C. (1993). Real world research: A resource for social sciences and practioner

researcher. Real world research: A resource for social sciences and practioner

researcher.

Rogers Everett, M. (1995). Diffusion of innovations. New York, 12.

Royal Philips Electronics. (2012). Philips Survey Reveals One in 10 Americans

Believe Online Health Information Saved Their Life. Retrieved August 9, 2015

from

http://www.newscenter.philips.com/us_en/standard/news/press/2012/

Ruane, J M (2005). The Essentials of Research Methods: A Guide to Social

Science Research. Wiley-Blackwell.

Rubin, A. M. (1981). An examination of television viewing

motivations. Communication Research--An International Quarterly, 8(2),

141-65.

Runkeeper. (2015). Retrieved August 9, 2015 from

http://runkeeper.com/

https://www.pwc.com/mx/es/industrias/archivo/2014-11-pwc-the-wearable-future.pdf

https://www.pwc.com/mx/es/industrias/archivo/2014-11-pwc-the-wearable-future.pdf

http://www.onebeacontech.com/OneBeaconTech/pages/news

http://www.newscenter.philips.com/us_en/standard/news/press/2012/20121212_Philips_Survey_Health_Info_Tech.wpd#.VcUzVfmqqkq

http://runkeeper.com/

56

RunKeeper. (2013). GPS Track Run Walk - Android Apps on Google Play. Retrieved

November 26, 2013 from

https://play.google.com/store/apps/details?id=com.fitnesskeeper.runkeeper.pro

Runtastic. (2015). Function. Retrieved August 9, 2015 from

https://www.runtastic.com/.

Saeed, K. A., Hwang, Y., & Mun, Y. Y. (2003). Toward an integrative framework for

online consumer behavior research: a meta-analysis approach. Journal of

Organizational and End User Computing (JOEUC), 15(4), 1-26.

Safavi, S., & Shukur, Z. (2014). Conceptual privacy framework for health information

on wearable device. PloS one, 9(12), e114306.

Salah, H., MacIntosh, E., & Rajakulendran, N. (2014). Wearable Tech: Leveraging

Canadian Innovation to Improve Health.

Samavi, R., Consens, M. P., & Chignell, M. (2014). PHR User Privacy Concerns and

Behaviours. Procedia Computer Science, 37, 517-524.

Straub, D. W. (1989). Validating instruments in MIS research. MIS quarterly,

147-169.

Straub, D., Boudreau, M. C., & Gefen, D. (2004). Validation guidelines for IS

positivist research. The Communications of the Association for Information

Systems, 13(1), 63.

Sultan, N. (2015). Reflective thoughts on the potential and challenges of wearable

technology for healthcare provision and medical education. International

Journal of Information Management, 35(5), 521-526.

Sun, N., & Rau, P. L. P. (2015). The acceptance of personal health devices among

patients with chronic conditions. International journal of medical informatics,

84(4), 288-297.

Tansey, O. (2007). Process tracing and elite interviewing: a case for non-probability

sampling. PS: Political Science & Politics, 40(04), 765-772.

Tehrani, K., & Michael, A. (2014). Wearable technology and wearable devices:

Everything you need to know. Wearable Devices Magazine.

https://play.google.com/store/apps/details?id=com.fitnesskeeper.runkeeper.pro

https://www.runtastic.com/

57

Tencent ISUX (2014). 2014 wearable devices white paper. Retrieved August 9, 2015

from

http://isux.tencent.com/wearable-devices-white-paper.html. In Chinese.

Tung, F. C., Chang, S. C., & Chou, C. M. (2008). An extension of trust and TAM

model with IDT in the adoption of the electronic logistics information system in

HIS in the medical industry. International journal of medical informatics, 77(5),

324-335.

Urbach, N., & Ahlemann, F. (2010). Structural equation modeling in information

systems research using partial least squares. JITTA: Journal of Information

Technology Theory and Application, 11(2), 5.

US Food and Drug Administration. (2013). Guidance for Industry. Patient-reported

outcome measures: use in medical product development to support labeling

claims. US Department of Health and Human Services Food and Drug

Administration, December 2009.

Varma Citrin, A., Sprott, D. E., Silverman, S. N., & Stem Jr, D. E. (2000). Adoption of

Internet shopping: the role of consumer innovativeness. Industrial management

& data systems, 100(7), 294-300.

Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the technology

acceptance model: Four longitudinal field studies. Management science, 46(2),

186-204.

Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of

information technology: Toward a unified view. MIS quarterly, 425-478.

Wahoo Fitness. (2015). Retrieved August 9, 2015 from

http://eu.wahoofitness.com/.

Warshaw, P. R. (1980). A New Model for Predicting Behavioral Intentions: An

Alternative to Fishbein, Journal of Marketing Research (17:2), pp. 153-172.

Wilson, E. V., & Lankton, N. K. (2004). Modeling patients' acceptance of

provider-delivered e-health. Journal of the American Medical Informatics

Association, 11(4), 241-248.

Wong, K. K. K. (2013). Partial least squares structural equation modeling (PLS-SEM)

techniques using SmartPLS. Marketing Bulletin, 24(1), 1-32.

Wright, R., & Keith, L. (2014). Wearable Technology: If the Tech Fits, Wear

It. Journal of Electronic Resources in Medical Libraries, 11(4), 204-216.

http://isux.tencent.com/wearable-devices-white-paper.html

http://eu.wahoofitness.com/

58

Wrightsman, L. S. (1991). Interpersonal trust and attitudes toward human nature.

Wu, J. H., & Wang, S. C. (2005). What drives mobile commerce? : An empirical

evaluation of the revised technology acceptance model. Information &

management, 42(5), 719-729.

Wu, L., & Chen, J. L. (2005). An extension of trust and TAM model with TPB in the

initial adoption of on-line tax: an empirical study. International Journal of

Human-Computer Studies, 62(6), 784-808.

Wu, J., Li, H., Lin, Z., & Goh, K. Y. (2015). How big data and analytics reshape the

wearable device market–the context of e-health. International Journal of

Production Research, 1-15.

59

Appendix. Questionnaire

Survey about the usage of smart bands in China

This survey is intended to study the usage patterns and opinions of people in China

about smart bands. The survey is part of an academic research project at University of

Oulu, Finland.

Smart bands are a kind of wearable devices that record and monitor real-time data

such as physical activity and sleeping patterns. They are typically used with mobile

applications (like in the picture).

Please answer the following questions as truthfully and accurately as possible. It takes

about 5 minutes to answer all questions. The participation to this survey is anonymous;

only summarized data will be reported in research. Your participation will help us

greatly. Thank You.

What type of mobile phone do you use in your everyday life?

○Basic mobile phone.

○Smartphone.

Rate the following statements using the 7-point scale:

1 2 3 4 5 6 7

Strongly

disagree

Moderately

disagree

Somewhat

disagree

Neutral Somewhat

agree

Moderately

agree

Strongly

agree

60

Technological personality (adapted from Koufaris, 2002)

MTS1. I am very skilled at using mobile technology.

MTS2. I know how to find what I want through mobile technology.

MTS3. I know more about using mobile technology than other users.

INO1. I think I know more about innovative digital products than my circle of friends

(Innovativeness, Aldaz et al. 2009)

INO2. I think I would use an innovative digital product even if nobody else I know

had used it before. (Innovativeness, Aldaz et al. 2009)

AFF1. Using a mobile phone is one of my main daily activities. (Affinity, Aldaz et

al.2009)

AFF2. If my mobile phone is down, I really miss it. (Affinity, Aldaz et al.2009)

AFF3. My mobile phone is important in my life. (Affinity, Aldaz et al.2009)

AFF4. I would be lost without my mobile phone. (Affinity, Aldaz et al.2009)

COM1. How often do you search for health related information on the Internet?

(Adapted Compatibility, Aldaz et al. 2009)

Never 1 2 3 4 5 6 7 Very often

COM2. How often do you use health applications on your mobile phone?

(Adapted Compatibility, Aldaz et al. 2009)

Never 1 2 3 4 5 6 7 Very often

Experience of using smart bands and its applications

Have you used smart bands before?

○Yes and still using now.

○Yes, but abandoned.

○No, but interested to.

○No, never wanted to use.

Please tell the name of smart bands you have used or currently use (select the one

used most frequently).

○Jawbone

○Fitbit

○Misfit

○Nike

○Adidas

○Apple

○Polar

○Others. Please specify………….

What type of applications have you used or do you currently use with smart

bands most frequently? (Please select all that apply)

□Sport activities.

□Sleep monitoring.

□Heart rate recording.

61

□Others. Please specify………….

For how long time have you used a smart band?

○Less than a week

○A few weeks, but less than a month

○1-2 months

○3-5 months

○6-11 months

○12-24 months

○Over two years

How often have you used or do you currently use the smart band?

○Daily

○Several times a week

○Every few weeks

○Once in few months or more rarely

○Never

Opinions about smart bands

Rate the following statements using the 7-point scale:

Perceived usefulness (adapted from Venkatesh and Davis, 2000)

PU1. Using the smart bands would help me monitor my physical health.

PU2. I think using smart bands would help me improve my physical health.

PU3. Using the smart bands would enhance my effectiveness in monitoring my

physical health.

PU4. Based on my perception of smart bands, I believe they provide good features.

Perceived ease of use (adapted from Venkatesh and Davis, 2000)

PEOU1. I think the interaction with the smart bands is clear and understandable.

PEOU2. It would be easy for me to become skilful at using smart bands.

PEOU3. I think it is easy to get the smart bands to do what I want it to do.

PEOU4. I think that it takes low mental effort to use smart bands.

Trust (adapted from Gefen et al., 2003)

TST1. I believe that the smart bands are reliable for the data recording.

TST2. I believe that the personal information is safe.

TST3. I believe that the smart bands and their applications provide accurate

information.

1 2 3 4 5 6 7

Strongly

disagree

Moderately

disagree

Somewhat

disagree

Neutral Somewhat

agree

Moderately

agree

Strongly

agree

62

Social influence (adapted from Chau & Hu PJH, 2002)

SI1. My relatives and friends think that I should use smart bands.

SI2. Product ads influence me to use smart bands.

SI3. Former users' comments influence me to use smart bands.

Behavioral intention to use smart bands (adapted from Venkatesh and Davis, 2000)

BI1. I intend to use smart bands in the future.

BI2. I predict I would use smart bands in the future.

Personality scales (adapted from Carver, C. S., & White, T. L, 1994)

Rate the following statements using scale:

1 2 3 4

very true for me somewhat true for me somewhat false for me very false for me

BIS: If I think something unpleasant is going to happen I usually get pretty "worked

up."

BAS Fun Seeking (BAFS): I'm always willing to try something new if I think it will

be fun.

BAS Reward Responsiveness (BARR): When I get something I want, I feel excited

and energized.

BAS Drive (BASD): When I want something I usually go all-out to get it.

Basic information

Gender

○Male ○Female

Country of residence

○China ○Other country…

Age

○under 18 ○18-24 ○25-29 ○30-34 ○35-39 ○40-44 ○45-49 ○50-54

○55-59 ○60-64 ○65 and over

Education

○Under high school ○High school ○College ○Graduate school (or above)

Monthly income

○¥ 0 – 4000 ○¥ 4001 – 8000 ○¥ 8001 – 12000 ○¥ 12001 +

Field of Work

○Sports

○IT

○Health

○Other …

the acceptance of wearable devices for personal...

Documents